Measurement properties of the Western Ontario Shoulder Instability Index in Dutch patients with shoulder instability
1 Department of Orthopaedic Surgery and Traumatology, Onze Lieve Vrouwe Gasthuis, Postbus 95500, 1090 HM Amsterdam, the Netherlands
2 Department of Orthopaedic Surgery and Traumatology, de Lairesse Kliniek, Amsterdam, the Netherlands
3 Department of Orthopaedic Surgery and Traumatology, Waterland Ziekenhuis, Purmerend, the Netherlands
4 Department of Epidemiology and Biostatistics and the EMGO Institute for Health and Care Research, VU Medical Center, Amsterdam, the Netherlands
BMC Musculoskeletal Disorders 2014, 15:211 doi:10.1186/1471-2474-15-211Published: 20 June 2014
The Western Ontario Shoulder Instability index (WOSI) is a patient-reported outcome measure for patients with shoulder instability. The purpose of this study was to validate the WOSI in a Dutch population by evaluating its structural validity, internal consistency, measurement error, reliability, and construct validity. Floor and ceiling effects were also addressed.
Two cohorts were recruited, including a total of 138 patients with shoulder instability. Confirmatory factor analysis was used to assess the structural validity and Cronbach’s α to assess internal consistency. The measurement error was calculated as the smallest detectable change (SDC). Reliability (test–retest) was estimated in a subgroup of 99 patients who completed the re-test after a mean of 13 days (5–30 days). Reliability was calculated with the intraclass correlation coefficient (ICC). Construct validity was evaluated by comparing the WOSI with the Oxford Shoulder Instability Score (OSIS), the Simple Shoulder Test, the Oxford Shoulder Score, the Disability of the Arm, Shoulder, and Hand assessment (DASH), and the Short Form-36 Health Survey. Measurement properties were evaluated for both the total WOSI score and its four domains.
Factor analysis did not confirm the validity of the four domains. Best results were found for a one-factor model. Internal consistency was good, with Cronbach’s α ranging from 0.93 to 0.96. Reliability was excellent (ICC 0.88–0.92 for all subscales). The measurement error (SDC) was 23.0% for the total WOSI and 23% to 28% for the subscales (on a scale of 0–100). Regarding the construct validity, 76% of the results were in accordance with the hypotheses, including a high correlation with the OSIS (0.82) and DASH (0.81) assessments. No floor or ceiling effects were found.
The Dutch version of WOSI showed good reliability and validity in a cohort of patients with shoulder instability, although the factor structure remains unclear.